Spatial variability of rainfall in the Colorado Rocky Mountains: A comparison of radar-derived and rain gauge data

Authors: SEARS, MEGAN - Colorado State University; KAMPF, STEPHANIE - Colorado State University; Barnard, David; SCHUMACHER, R - Colorado State University; MILLER, QUINN - Colorado State University; RICHARDSON, MIKAELA - Colorado State University; WEDER LOYA, SHELBY - Colorado State University; STRUTHERS, SAMUEL - Colorado State University; RASMUSSEN, K - Colorado State University

Submitted to: Journal of Hydrology: Regional Studies
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/10/2026
Publication Date: 8/15/2026
Citation: Sears, M.G., Kampf, S., Barnard, D.M., Schumacher, R., Miller, Q., Richardson, M., Weder Loya, S., Struthers, S., Rasmussen, K.L. 2026. Spatial variability of rainfall in the Colorado Rocky Mountains: A comparison of radar-derived and rain gauge data. Journal of Hydrology: Regional Studies. 67. Article 103858. https://doi.org/10.1016/j.ejrh.2026.103858.
DOI: https://doi.org/10.1016/j.ejrh.2026.103858

Interpretive Summary: Understanding how intense rainfall varies across mountain landscapes is important for predicting flash floods and debris flows, especially after wildfires. This study looked at rainfall patterns in the northern Colorado Rocky Mountains and evaluated how well different types of rainfall data predict streamflow in burned areas. Results indicate that the east side of the mountains generally experienced more total rainfall and more intense storms than the west side, likely due to regional weather patterns that favor storm development on the plains side. However, there was no consistent link between elevation and rainfall intensity, as storm patterns shifted from year to year. Three rainfall datasets were compared: tipping bucket gauges on the ground, radar-based MRMS data, and the gridded gridMET product. Radar data performed better for predicting streamflow on the east side, where radar coverage is strong, while tipping buckets were more accurate on the west side, where radar signals are weaker due to mountain terrain. These results highlight the need to choose rainfall data carefully, especially in post-fire environments where accurate rain measurements are critical for managing flood risks.

Technical Abstract: Flash floods in the Colorado Rockies are more likely after intense rainstorms, especially in areas affected by wildfires. This study investigates the spatial variability of convective rainfall across the northern Colorado Rocky Mountains, with a focus on how rainfall intensity relates to elevation and how different rainfall datasets perform in predicting streamflow response in post-wildfire landscapes. Using three summers of data from tipping bucket gauges, the radar-derived Multi-Radar/Multi-Sensor (MRMS) product, and the gridMET dataset, we found that rainfall totals and intensities were generally higher on the eastern side of the Rocky Mountain divide, likely due to regional convective processes such as the mountain–plains solenoidal circulation. However, no consistent relationship between elevation and rainfall intensity emerged across years, reflecting the complexity and spatial heterogeneity of convective storm dynamics. Comparisons between rainfall datasets revealed substantial differences: MRMS estimated more rainfall east of the mountains but less to the west compared to gridMET, likely due to differences in radar quality and gauge density. At the storm-event scale, MRMS and tipping buckets were moderately correlated, but accuracy varied by location. MRMS performed better in predicting streamflow response east of the divide while tipping buckets were more effective in areas with poor radar coverage. Overall, while dense tipping bucket networks remain the most accurate for capturing localized convective rainfall, MRMS offers a viable alternative in radar-optimized regions and can better represent spatial variability critical for hydrologic hazard prediction in data-scarce, fire-affected watersheds.